Open to opportunities

Bhavana Maganti

@bhavanamaganti

Message

I build and evaluate multi-agent AI systems, retrieval workflows, and validated coding solutions.

United States

Message

What I'm looking for

I'm looking to build and evaluate reliable AI systems, contribute to rigorous model validation, and collaborate remotely on well-documented open-source and software development work.

At the National Institute for Aviation Research, I design a Python-based multi-agent orchestration layer to evaluate and improve AI model performance.

I develop and review real-world Python and C++ coding tasks for fault-triage scenarios, helping improve model accuracy and validation consistency. I validate solutions with competitive-programming techniques and maintain output consistency above 95%.

I built DocuChat, a retrieval-augmented chatbot over PDFs using recursive character chunking, sentence-transformer embeddings, and ChromaDB. I also implemented ResearchCrew, a LangGraph-based multi-agent research assistant for collaborative research and summarization workflows.

I document model behavior, prompt templates, security constraints, and implementation decisions clearly. I enjoy rigorous code validation, open-source collaboration, and building dependable AI tooling for remote technical teams.

Experience

Work history, roles, and key accomplishments

Current

AI Engineer

Current

National Institute for Aviation Research (NIAR)

May 2025 - Present (1 year 4 months)

Designed a multi-agent orchestration layer in Python3 to evaluate and improve AI model performance. Developed and reviewed coding tasks in Python3 and C++ to analyze fault-triage scenarios, improving model accuracy and validation consistency.